CN112800867B - Pine wood nematode disease dead tree detection method based on two-stage high-altitude tripod head video - Google Patents
Pine wood nematode disease dead tree detection method based on two-stage high-altitude tripod head video Download PDFInfo
- Publication number
- CN112800867B CN112800867B CN202110040676.3A CN202110040676A CN112800867B CN 112800867 B CN112800867 B CN 112800867B CN 202110040676 A CN202110040676 A CN 202110040676A CN 112800867 B CN112800867 B CN 112800867B
- Authority
- CN
- China
- Prior art keywords
- value
- image
- altitude
- pine wood
- stage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
Abstract
The invention discloses a pine wood nematode disease dead tree detection method based on a two-stage high-altitude tripod head video, which comprises a first stage and a second stage, wherein the first stage directly and rapidly positions the PTZ value of the occurrence position of a suspicious disease dead tree by using the video; and in the second stage, the suspicious disease dead tree is photographed again, and further identification processing is carried out by adopting a pine wood nematode disease dead tree CNN classification algorithm. According to the method, firstly, the disease and dead tree is rapidly positioned through the high-altitude holder video, and then the pine wood nematode disease and dead tree are classified and identified through shooting again, so that the process of identifying the dead and dead tree of the high-altitude flame-proof holder for forestry can be greatly accelerated, and the accuracy of an identification result is further improved; the device does not need to add any new equipment at the front end, and can cooperate with the function of the high-altitude fire-prevention cradle head.
Description
Technical Field
The invention relates to the technical field of forest protection, in particular to a pine wood nematode disease dead tree detection method based on a two-stage high-altitude tripod head video.
Background
Pine wood nematodes belong to the phylum of the Linear animal, order Apocynaceae, family Apocynaceae, genus Apocynaceae. Is one of foreign invasive species with larger harm in China (but is not listed in the first foreign invasive species list in China). In 1982, it was first discovered in Nanjing Zhongshanling that several centers of diseases were formed in Anhui, shandong, zhejiang, guangdong and the like successively, and spread all around, causing local areas of these provinces to undergo concurrent disaster, resulting in death of a large number of pine trees. The economic loss caused by pine wood nematode disease to Anhui and Zhejiang provinces is up to 5-7 hundred million yuan.
Because of the destructive hazard of pine nematodes, the worms have been listed as important quarantine targets for both the inside and outside. Close range propagation is mainly carried and propagated by a medium longicorn such as Monochamus alternatus (Monochamus alternatus); the propagation is mainly carried out by manually transferring seedlings, pine wood packaging boxes, pine wood products and the like with epidemic disease (longicorn with pine wood nematodes). Pine tree infected by pine wood nematode has needle with yellow brown or reddish brown, wilting, and stopped resin secretion, and longhorn penetration hole or spawning trace, and the whole tree is dead and blue. Seriously threatening the forests. Because of rapid expansion, the natural conifer forests in various scenic spots such as Huangshan and Zhangjiu have become a great threat.
Once the diseased plants are found, the diseased plants need to be cut down, so that the spread of pine wood nematode diseases is prevented, and the aims of reducing economic loss and protecting ecology are fulfilled. At present, the method for detecting and treating pine wood nematode disease in China is mainly artificial patrol, but the aim of early treatment is difficult to achieve due to the rare course of people in many natural protection areas.
The defects of the prior art are that: some research institutions use unmanned aerial vehicles to patrol a forest farm and use corresponding algorithms to identify the forest farm, but the unmanned aerial vehicles have the technical problem of limited patrol range, and the unmanned aerial vehicles need to participate manually in the whole process; the existing high-altitude holder is used for detecting the dead pine wood nematode disease, the detection is not realized through video, but is realized through fixed-point photographing, and the method has the advantages of long time consumption and possibility of affecting the normal fireproof target of the high-altitude holder.
Therefore, the invention hopes to further utilize the existing high-altitude holder to achieve a method for positioning the dead tree of pine wood nematodes in real time, efficiently and more accurately.
Related data:
1. pine wood nematodes, see hundred degrees encyclopedia:
https://baike.baidu.com/item/％E6％9D％BE％E6％9D％90％E7％BA％BF％E8％99％AB；
2. laplace operator:
http://www.opencv.org.cn/opencvdoc/2.3.2/html/doc/tutorials/imgproc/imgtrans/laplace_operator/laplace_operator.html；
3. the chinese secret algorithm:
https://github.com/ouyanghuiyu/chineseocr_lite；
4、Yolo v3，https://github.com/pjreddie/darknet；
5: VGG model:
https://baike.baidu.com/item/VGG％20％E6％A8％A1％E5％9E％8B/22689655。
disclosure of Invention
In view of at least one defect in the prior art, the invention aims to provide a pine wood nematode disease dead tree detection method based on a two-stage high-altitude tripod head video, which comprises the steps of rapidly positioning a disease dead tree through the high-altitude tripod head video, and then carrying out classification and identification on the pine wood nematode disease dead tree through shooting again, so that the process of identifying the dead tree through a forestry high-altitude flame-proof tripod head can be quickened, and the accuracy of an identification result is further improved; the device does not need to add any new equipment at the front end, and can cooperate with the function of the high-altitude fire-prevention cradle head.
In order to achieve the above purpose, the invention adopts the following technical scheme: a pine wood nematode disease dead tree detection method based on a two-stage high-altitude tripod head video is characterized by comprising a first stage and a second stage, wherein the first stage is used for directly and quickly positioning a PTZ value of an appearance position of a suspicious disease dead tree by using the video; and in the second stage, the suspicious disease dead tree is photographed again, and further identification processing is carried out by adopting a pine wood nematode disease dead tree CNN classification algorithm.
According to the detection method, the disease and dead tree is rapidly located through the video of the high-altitude tripod head camera, and then classified and identified through shooting again.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized in that; the method comprises the following steps:
step one: generating a PTZ scanning information video which rotates horizontally by utilizing a parameter value Z of a high-altitude pan-tilt camera;
step two: for each frame image F of the video generated in step one k Performing recognition pretreatment; removing the blurred image;
step three: performing target detection and identification on the pretreatment result of the step two by adopting a target detection pine wood nematode algorithm, and extracting suspicious image results;
step four: the image F needing repositioning shooting in the third step k Reprocessing, calling a high-altitude pan-tilt camera to zoom and position, and photographing to generate a new positioning image; performing recognition pretreatment of the second stage; then, target detection and identification are carried out on the pine wood nematode through a target detection algorithm;
step five: and carrying out classification identification by using a classified pine wood nematode dead tree CNN classification algorithm, and finding out an image of a classification result.
The method comprises the steps of preprocessing through identification; removing the blurred image; target detection and identification are carried out on the pretreatment result by adopting a target detection pine wood nematode algorithm, and suspicious image results are extracted; and then, classifying and identifying by using a classified pine wood nematode dead tree CNN classification algorithm, and finding out an image of a classification result. Thereby finding out the dead tree image of the pine wood nematode disease.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized in that; the first step comprises the following steps:
step 1.1: turning on PTZ display function of high altitude PTZ camera, and simultaneously reading zoom multiple parameter of high altitude PTZ camera, namely parameter value Z, and obtaining minimum value Z of parameter value Z min Maximum Z max ；
Step 1.2: triple-loop scanning is carried out on PTZ of the high-altitude PTZ camera, video is shot through the high-altitude PTZ camera, and the P value is from P min ～P Max Regulation, P min Is the minimum value of P value, P Max Is the maximum value of the P value; t value from T min ～T max Regulation, T min Is the minimum value of T max Is the minimum of the T value; z value from Z min To Z max Adjusting; the innermost loop is the P-value loop, whenever it reaches a minimum value of P min Or maximum value P Max When the T value is regulated; when the T value reaches the maximum value T max Or minimum value T min When, then the Z value is adjusted.
According to the method, triple cyclic scanning is carried out through the PTZ of the high-altitude PTZ camera, and video is shot through the high-altitude PTZ camera.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized in that; the step 1.2 comprises the following steps:
step 1.2.1: the high-altitude pan-tilt camera moves horizontally at a speed which is 1/m times the maximum horizontal movement speed, and m is more than 1; horizontally scanning and shooting video;
step 1.2.2: if the P value is greater than P min And is less than P max Returning to the step 1.2.1; when the P value is greater than or equal to P max Or less than or equal to P min When the horizontal rotation is reversed, the T value is regulated by the stepping value delta T, and delta T represents the stepping value of the high-altitude tripod head camera in the vertical direction; step 1.2.3;
step 1.2.3: if the T value is greater than T min Less than T max Returning to the step 1.2.1; if T value>T max Then T value = T max The method comprises the steps of carrying out a first treatment on the surface of the If T value<T min Then T value = T min The method comprises the steps of carrying out a first treatment on the surface of the Adjusting the Z value by delta Z, wherein delta Z represents the adjustment multiple of each scaling; turning to step 1.2.4;
step 1.2.4: if the Z value is less than Z max Returning to the step 1.2.1; if Z value is greater than or equal to Z max The first phase scan ends.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the second step comprises the following steps:
step 2.1: sequentially reading each frame image F in the video of Gao Kongyun cameras k ；
Step 2.2: judging image F using OpenCV algorithm k Whether the images belong to a fuzzy image; if so, discard the image F k Turning to step 2.1; otherwise, turning to step 2.3;
step 2.3: for image F k OCR recognition is carried out to obtain an image F k In real time PTZ value, P k T k Z k Image F k And P k T k Z k The values are combined together.
According to the method, the OpenCV algorithm is adopted to remove the fuzzy image; and image F k And P k T k Z k The values are combined together.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the step 2.2 includes: for image F k Performing Laplacian transformation, judging its return value, taking 100 as threshold value, if the return value is less than 100, the image F k Discarding the image F as a blurred image k Turning to step 2.1; if the return value is greater than 100, the clear diagram is considered, and the step 2.3 is shifted.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the third step comprises the following steps:
step 3.1: reading an image F k Image data of (2) and corresponding PTZ value, i.e. P k T k Z k Putting the target into a target pine wood nematode detection algorithm for identification, wherein the target pine wood nematode detection algorithm is a trained YOLO V3 algorithm, and identifying whether the target pine wood nematode detection algorithm is an image of dead trees or not to obtain an identification result;
then, selecting the identification result coordinates by a frame, and outputting the identification result coordinates as follows:
aBoxs＝[{x 11 ,y 11 ,x 12 ,y 12 },{x 21 ,y 21 ,x 22 ,y 22 },...,{x i1 ,y i1 ,x i2 ,y i2 },...]the method comprises the steps of carrying out a first treatment on the surface of the Wherein x is i1 ，y i1 Representation of
The upper left corner coordinate, x, corresponding to the identified dead tree image i i2 ,y i2 Representing its lower right angular position;
step 3.2: judging the current image F k Is an image of dead trees? If not, discard the image F k Go to step 3.1; if yes, turning to step 3.3;
step 3.3: judging { x } i1 ,y i1 ,x i2 ,y i2 Whether or not the framed image area is smaller than image F k 1/4 of the picture of (a), if so, image F k Marked with { x } i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k 1, need to relocate the photo; otherwise, image F k Marking, namely: { x i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k 0, without repositioning the photograph, put the marked result into the mark array R1;
step 3.4: image F to be photographed without repositioning k According to { x } i1 ,y i1 ,x i2 ,y i2 Cutting the coordinate position, and putting the cutting result into an array R2;
step 3.5: and judging whether all the videos generated in the step two are processed, if yes, turning to the step four, otherwise turning to the step 3.1.
According to the method, the image of the dead tree is identified through the target pine wood nematode detection algorithm. The image i of the dead tree refers to the image of the ith dead tree.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the fourth step comprises the following steps:
step 4.1: sequentially reading the position information of the image information in the array R1, namely the array generated in the step 3.3
aBoxs＝[{x 11 ,y 11 ,x 12 ,y 12, P 1 ,T 1 ,Z 1 ,x 1 },{x 21 ,y 21 ,x 22 ,y 22, P 2 ,T 2 ,Z 2 ,x 2 },,...,{x i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k ,x i },...]；
Step 4.2: if { x } i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k ,x i X in } i =1, then according to the resolution and x of the current video of the aerial pan-tilt camera i1 ,y i1 ,x i2 ,y i2 Calling a 3D positioning API of the high-altitude tripod head camera according to the coordinate x i1 -10,y i1 -10,x i2 +10,y i2 +10, scaling and positioning;
step 4.3: photographing to generate a new positioning image, carrying out target detection and identification on the new positioning image, and then, selecting the identification result coordinate { x 'again through a target detection pine wood nematode algorithm' i1 ,y’ i1 ,x’ i2 ,y’ i2 }；
Step 4.4: if the identification result is not the image of dead trees, discarding the image, and turning to the step 4.1; otherwise, the PTZ value is read again, and the output result is { x' i1 ,y’ i1 ,x’ i2 ,y’ i2, P’ k ,T’ k ,Z’ k ,0}；
Step 4.5 according to the new recognition result coordinates { x' i1 ,y’ i1 ,x’ i2 ,y’ i2 The recognition result is cut out and put into an array R2.
According to the method, the 3D positioning API of the high-altitude tripod head camera is called for the image needing to be re-photographed according to the coordinate x i1 -10,y i1 -10,x i2 +10,y i2 +10, scaling and positioning; and photographing to generate a new positioning image, and carrying out target detection and identification on the new positioning image.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the fifth step comprises the following steps:
step 5.1: sequentially reading image information F 'in array R2' i ；
Step 5.2: for image information F' i Classifying and identifying by using a classified pine wood nematode disease dead tree CNN classification algorithm, wherein the method uses a trained VGG classification algorithm;
step 5.3: if so, outputting image information F' i A corresponding PTZ value; if not, delete image information F' i Turning to step 5.4;
step 5.4: determine if array R2 has been processed? If not, turning to step 5.1, if yes, ending.
According to the method, classification and identification are carried out through a two-classification dead pine nematode tree CNN classification algorithm, and dead pine nematode trees are identified.
The invention provides a pine wood nematode dead tree detection method based on a two-stage high-altitude tripod head video, which comprises the steps of rapidly positioning the pine wood nematode dead tree through the high-altitude tripod head video, and then carrying out classification and identification on the pine wood nematode dead tree through shooting again, so that the process of identifying the pine wood nematode dead tree through the high-altitude flame-proof tripod head in forestry can be quickened, and the accuracy of an identification result is further improved; the device does not need to add any new equipment at the front end, and can cooperate with the function of the high-altitude fire-prevention cradle head.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a detailed flow chart of the present invention;
FIG. 3 is a first stage flow chart of the present invention;
fig. 4 is a second stage flow chart of the present invention.
Detailed Description
The invention will be described in further detail with reference to the drawings and the specific examples.
1-4, the pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video comprises a first stage and a second stage, wherein the first stage is used for rapidly positioning the PTZ value of the occurrence position of a suspicious disease dead tree directly by using the video; and in the second stage, the suspicious disease dead tree is photographed again, and further identification processing is carried out by adopting a pine wood nematode disease dead tree CNN classification algorithm.
The first stage identifies suspicious disease and the second stage identifies disease caused by pine wood nematode disease.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized in that; the method comprises the following steps:
the first stage comprises a first step, a second step and a third step, and the second stage comprises a fourth step and a fifth step;
step one: generating a PTZ scanning information video which rotates horizontally by utilizing a parameter value Z of a high-altitude pan-tilt camera;
step two: for each frame image F of the video generated in step one k Performing recognition pretreatment; removing the blurred image;
step three: performing target detection and identification on the pretreatment result of the step two by adopting a target detection pine wood nematode algorithm, and extracting suspicious image results; finding out suspicious disease withered trees;
step four: the image F needing repositioning shooting in the third step k Reprocessing, calling a high-altitude pan-tilt camera to zoom and position, and photographing to generate a new positioning image; performing recognition pretreatment of the second stage; then, target detection and identification are carried out on the pine wood nematode through a target detection algorithm;
step five: and carrying out classification identification by using a classified pine wood nematode dead tree CNN classification algorithm, and finding out an image of a classification result.
The high-altitude pan-tilt camera is the existing equipment.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized in that; the first step comprises the following steps:
step 1.1: turning on PTZ display function of high altitude PTZ camera, and simultaneously reading zoom multiple parameter of high altitude PTZ camera, namely parameter value Z, and obtaining minimum value Z of parameter value Z min Maximum Z max ；
Step 1.2: triple-loop scanning is carried out on PTZ of the high-altitude PTZ camera, video is shot through the high-altitude PTZ camera, and the P value is from P min ～P Max Regulation, P min Is the minimum value of P value, P Max Is the maximum value of the P value; t value from T min ～T max Regulation, T min Is the minimum value of T max Is the minimum of the T value; z value from Z min To Z max Adjusting; the innermost loop is the P-value loop, whenever it reaches a minimum value of P min Or maximum value P Max When the T value is regulated; when the T value reaches the maximum value T max Or minimum value T min When, then the Z value is adjusted.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized in that; the step 1.2 comprises the following steps:
step 1.2.1: the high-altitude pan-tilt camera moves horizontally at a speed which is 1/m times the maximum horizontal movement speed, and m is more than 1; horizontally scanning and shooting video; during scanning, the video obtained by the high-altitude tripod head camera can be processed in the second step and the third step; the speed is improved;
step 1.2.2: if the P value is greater than P min And is less than P max Returning to the step 1.2.1; when the P value is greater than or equal to P max Or less than or equal to P min When the horizontal rotation is reversed, the T value is regulated by the stepping value delta T, and delta T represents the stepping value of the high-altitude tripod head camera in the vertical direction; step 1.2.3;
step 1.2.3: if the T value is greater than T min Less than T max Return toStep 1.2.1; if T value>T max Then T value = T max The method comprises the steps of carrying out a first treatment on the surface of the If T value<T min Then T value = T min The method comprises the steps of carrying out a first treatment on the surface of the Adjusting the Z value by delta Z, wherein delta Z represents the adjustment multiple of each scaling; turning to step 1.2.4;
step 1.2.4: if the Z value is less than Z max Returning to the step 1.2.1; if Z value is greater than or equal to Z max The first phase scan ends.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the second step comprises the following steps:
step 2.1: sequentially reading each frame image F in the video of Gao Kongyun cameras k ；
Step 2.2: judging image F using OpenCV algorithm k Whether the images belong to a fuzzy image; if so, discard the image F k Turning to step 2.1; otherwise, turning to step 2.3;
step 2.3: for image F k OCR recognition is carried out to obtain an image F k In real time PTZ value, P k T k Z k Image F k And P k T k Z k The values are combined together.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the step 2.2 includes: for image F k Performing Laplacian transformation, judging its return value, taking 100 as threshold value, if the return value is less than 100, the image F k Discarding the image F as a blurred image k Turning to step 2.1; if the return value is greater than 100, the clear diagram is considered, and the step 2.3 is shifted.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the third step comprises the following steps:
step 3.1: reading an image F k Image data of (2) and corresponding PTZ value, i.e. P k T k Z k Putting the target into a target pine wood nematode detection algorithm for identification, wherein the target pine wood nematode detection algorithm is a trained YOLO V3 algorithm, and identifying whether the target pine wood nematode detection algorithm is dead trees or notObtaining an identification result;
then, selecting the identification result coordinates by a frame, and outputting the identification result coordinates as follows:
aBoxs＝[{x 11 ,y 11 ,x 12 ,y 12 },{x 21 ,y 21 ,x 22 ,y 22 },...,{x i1 ,y i1 ,x i2 ,y i2 },...]the method comprises the steps of carrying out a first treatment on the surface of the Wherein x is i1 ，y i1 Representation of
The upper left corner coordinate, x, corresponding to the identified dead tree image i i2 ,y i2 Representing its lower right angular position;
step 3.2: judging the current image F k Is an image of dead trees? If not, discard the image F k Go to step 3.1; if yes, turning to step 3.3;
step 3.3: judging { x } i1 ,y i1 ,x i2 ,y i2 Whether or not the framed image area is smaller than image F k 1/4 of the picture of (a), if so, image F k Marked with { x } i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k 1, need to relocate the photo; otherwise, image F k Marking, namely: { x i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k 0, without repositioning the photograph, put the marked result into the mark array R1;
step 3.4: image F to be photographed without repositioning k According to { x } i1 ,y i1 ,x i2 ,y i2 Cutting the coordinate position, and putting the cutting result into an array R2;
step 3.5: and judging whether all the videos generated in the step two are processed, if yes, turning to the step four, otherwise turning to the step 3.1.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the fourth step comprises the following steps:
step 4.1: sequentially reading the position information of the image information in the array R1, namely the array generated in the step 3.3
aBoxs＝[{x 11 ,y 11 ,x 12 ,y 12, P 1 ,T 1 ,Z 1 ,x 1 },{x 21 ,y 21 ,x 22 ,y 22, P 2 ,T 2 ,Z 2 ,x 2 },,...,{x i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k ,x i },...]；
Step 4.2: if { x } i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k ,x i X in } i =1, then according to the resolution and x of the current video of the aerial pan-tilt camera i1 ,y i1 ,x i2 ,y i2 Calling a 3D positioning API of the high-altitude tripod head camera according to the coordinate x i1 -10,y i1 -10,x i2 +10,y i2 +10, scaling and positioning;
step 4.3: photographing to generate a new positioning image, carrying out target detection and identification on the new positioning image, and then, selecting the identification result coordinate { x 'again through a target detection pine wood nematode algorithm' i1 ,y’ i1 ,x’ i2 ,y’ i2 }；
Step 4.4: if the identification result is not the image of dead trees, discarding the image, and turning to the step 4.1; otherwise, the PTZ value is read again, and the output result is { x' i1 ,y’ i1 ,x’ i2 ,y’ i2, P’ k ,T’ k ,Z’ k ,0}；
Step 4.5 according to the new recognition result coordinates { x' i1 ,y’ i1 ,x’ i2 ,y’ i2 The recognition result is cut out and put into an array R2.
The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video is characterized by comprising the following steps of: the fifth step comprises the following steps:
step 5.1: sequentially reading image information F 'in array R2' i ；
Step 5.2: for image information F' i Death of pine wood nematodes using two classificationsThe tree CNN classification algorithm carries out classification identification, and the method uses a training VGG classification algorithm;
step 5.3: if so, outputting image information F' i A corresponding PTZ value; if not, delete image information F' i Turning to step 5.4;
step 5.4: determine if array R2 has been processed? If not, turning to step 5.1, if yes, ending.
The PTZ security monitoring equipment concept Pan/Tilt/Zoom is abbreviated and represents the omnibearing (left-right/up-down) movement of a tripod head and Zoom control of a lens. The term is generally mentioned and can be understood colloquially as a pan-tilt control.
In the scene range monitored by the camera, after a moving target appears, a user can manually lock (for example, the target is locked by clicking a mouse) or a preset position is automatically triggered to lock a certain moving target, so as to trigger the PTZ camera to carry out autonomous and automatic PTZ tracking, automatically control a cradle head of the PTZ camera to carry out omnibearing rotation, and carry out vision-guided automatic tracking on the locked moving target, thereby ensuring that the tracking target continuously appears in the center of a lens. The automatic PTZ tracking module overcomes the defect of narrow monitoring field of the fixed camera, and is a necessary function of a perfect safety monitoring system.
The detection method is described in further detail below with reference to the accompanying drawings.
Pine wood nematode disease dead tree detection method based on two-stage high-altitude holder video identification.
As shown in fig. 1, it uses two different phases: the first stage adopts PTZ values for directly and rapidly locating the occurrence position of the suspicious disease tree in the video, and the second stage takes a new picture of the suspicious disease tree, and adopts a classification algorithm of the dead and dead pine of the pine wood nematode to carry out further identification processing.
As shown in fig. 2-4, further comprising the steps of:
step one: directly utilizing the Z parameter value of the existing high-altitude tripod head camera to generate PTZ scanning information of horizontal rotation;
it comprises the following steps:
step 1.1 turning on the PTZ display function of the camera while reading the zoom factor parameter of the camera, i.e. Z max Values. Examples: since the Z value is at least 1 time, there are: z is Z min =1 to Z max =50 times.
Step 1.2 triple cycle scanning of PTZ, wherein the P value is from P min ～P Max I.e. 0 to 360 degrees, T being from T min ～T max The method comprises the steps of carrying out a first treatment on the surface of the Namely-90 to 0 DEG, Z value is from Z min To Z max Associated with the camera, e.g. 1-50 times. To avoid excessive view overlapping, the scan is completed as soon as possible, the innermost layer is cycled to a P value, rotated horizontally, and increased by a T value whenever it reaches a minimum or maximum; when the T value reaches the maximum value T max Or minimum value T min When the Z value is increased:
step 1.2 comprises the following sub-steps:
step 1.2.1: let Δt=10, Δz=z max And/10, which represent the step value in the vertical direction and the increase factor per scaling, respectively. Starting from p=0, t= -80, z=1, the camera moves horizontally at 1/5 times the maximum speed.
Step 1.2.2 the P value will increase more slowly from 0 to 360 degrees or decrease from 360 degrees to 0 degrees due to the direct horizontal scan. At the same time as the scan, video F is processed, see: and step two, performing step two.
When the P value is greater than or equal to P max Or less than or equal to P min When the horizontal rotation is reversed, namely: the rightward movement is changed to leftward movement, and the leftward movement is changed to rightward movement. Meanwhile, t=t+Δt.
Step 1.2.3: if a new T value>T end Or (b)<T start Then T value = T start Or T end And z=z+Δz, while Δt is inverted, i.e.: if the upward movement is changed to downward movement, the downward movement is changed to upward movement.
Step 1.2.4 if the Z value is between 1-10 times, Δt=10; if the Z value is between 10-30 times, Δt=5; if the Z value is greater than Z max The first phase scan ends. Otherwise, turning to 1.2.2.
Step two: because the step is to continuously generate the video for a while, the step two is to carry out the first stage identification pretreatment on each frame;
it comprises the following substeps:
step 2.1 reading each frame image F in the video k 。
Step 2.2, since it is not known about the environment, faces the problem of continuously capturing blurred images, such as: the camera is continually enlarged against the camera mount, or against nearby objects, and thus all blurred images need to be removed. We use OpenCV to make fuzzy class decisions by: the image is subjected to Laplacian transformation, the return value is judged, 100 is taken as a threshold value, if the return value is smaller than 100, the image is discarded, and the step 2.1 is carried out. If the number is greater than 100, the clear diagram is considered, and the next step is performed.
Step 2.3 it is difficult to read the correct PTZ value directly from the pan-tilt, since the pan-tilt is constantly moving. However, since the PTZ display switch is turned on in step 1.1, the camera displays the PTZ value in the video in real time, and thus OCR (optical character recognition) is performed on it. The method uses chineseocerlite to perform OCR recognition of a designated area, and the result is a real-time PTZ value, F k And P k T k Z k The values are combined together and the process proceeds to the next step;
step three: f to be identified k Performing target detection and identification on the image in the first stage, and extracting suspicious image results;
step 3.1 reading a strip F k Is noted as: PTZ is put into a trained target detection pine wood nematode algorithm for identification. The target detection algorithm selects a trained YOLO V3 algorithm which automatically identifies dead trees, and then frames the identification result, and outputs the result as abox= [ { x 11 ,y 11 ,x 12 ,y 12 },{x 21 ,y 21 ,x 22 ,y 22 },...,{x i1 ,y i1 ,x i2 ,y i2 },...]. Wherein x is i1 ，y i1 Representing the upper left corner coordinate, x, corresponding to the identified tree i i2 ,y i2 Indicating its lower right angular position.
Step 3.2: is the current image identified as a dead tree? If not, the image is discarded and the process goes to step 3.1.
Step 3.3 Loop determination { x } i1 ,y i1 ,x i2 ,y i2 Whether or not the value of } is less than 1/4 of the picture, i.e. determining (x) i2 -x i1 )×(y i2 -y i1 ) Whether it is smaller than 1/4 of the picture size; it is determined whether the identified box is large enough and the identification is not re-enlarged. If so, the element is extended, marked { x } i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k 1}, which represents: repositioning the photograph is required; otherwise, it is noted that: { x i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k 0, which represents: the result after marking is put into the marking array R1 without repositioning the photographing.
Step 3.4 the photographed image is not required to be repositioned according to x i1 ,y i1 ,x i2 ,y i2 And (5) cutting the position, and putting the cutting result into an array R2.
Step 3.5 unless the video generated in step two is completely processed, the process goes to step four, otherwise, the process goes to step 3.1.
Step four: reprocessing the image needing repositioning photographing in the third step; at this point, since the first-stage scanning of the camera has been completed, a second-stage recognition pre-process can be performed, which comprises the following sub-steps:
step 4.1, sequentially reading the position information in the image information in the array R1, namely, the array generated in step 3.4:
aBoxs＝[{x 11 ,y 11 ,x 12 ,y 12, P 1 ,T 1 ,Z 1 ,x 1 },{x 21 ,y 21 ,x 22 ,y 22, P 2 ,T 2 ,Z 2 ,x 2 },,...,{x i1 ,y i1 ,x i2 ,y i2,
P k ,T k ,Z k ,x i },...]the method comprises the steps of carrying out a first treatment on the surface of the Wherein x is 0 or 1.
Step 4.2 if element i= { x i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k ,x i X in } i =1 indicates that the photograph needs to be repositioned, then according to the resolution (e.g., 1920×1080) and x of the current video i1 ,y i1 ,x i2 ,y i2 Calling the 3D positioning API of the camera according to the coordinate x i1 -10,y i1 -10,x i2 +10,y i2 +10 performs zoom positioning. This step ensures that a larger region of interest is generated.
Step 4.3, photographing to generate a new positioning image, repeatedly performing target detection and identification on the new positioning image, and outputting a reclassified identification result { x 'by the algorithm' i1 ,y’ i1 ,x’ i2 ,y’ i2 }。
And 4.4, if the identification result is not a disease tree, turning to the step 4.1. Otherwise, the PTZ value at that time is read again, and the output at that time is updated to { x' i1 ,y’ i1 ,x’ i2 ,y’ i2, P’ k ,T’ k ,Z’ k ,0}。
Step 4.5, according to the new identification coordinates: x's' i1 ,y’ i1 ,x’ i2 ,y’ i2 Cutting out the identification result and putting the identification result into an R2 array.
Step five: carrying out second recognition on the image in the first stage to find out an image of the classification result; the classification can greatly improve the recognition precision from 40% to more than 80%, and the difficulty of low target detection precision is truly solved;
the sub-steps are as follows:
step 5.1 sequentially reading the image information F 'in the array R2' i ；
And 5.2, classifying and identifying the images by using a classified pine wood nematode dead tree CNN classification algorithm, wherein the method uses a VGG classification algorithm trained by the user.
Step 5.3 if yes, outputting F' i And corresponding PTZ values.
Step 5.4 is array R2 processed? If not, turning to step 5.1, otherwise, ending the process.
The method realizes the whole process of pine wood nematode disease dead tree detection based on two-stage high-altitude holder video recognition, and the flow charts of two corresponding different stages are shown in fig. 3-4.
The key points of the technology are as follows:
comprising the following steps: firstly, directly utilizing a basic Z value of an existing high-altitude tripod head camera to perform triple-cycle video scanning of PTZ; and taking each frame in the video image, carrying out the first-stage disease tree target detection and identification on the image to be identified, and extracting suspicious image results. Then, preprocessing the image of the smaller image, which needs to be relocated for photographing, in the second stage, relocating for photographing, and finding out the larger suspicious image again; finally, in order to improve the accuracy of the pine wood nematodes, classifying the cut images by using a classification algorithm of the pine wood nematode disease withered tree, and outputting the final classification result.
Technical protection scope abstract:
a pine wood nematode disease dead tree detection and positioning method based on high-altitude holder identification is characterized by comprising the following steps:
firstly, all rotating and amplifying PTZ track information is directly generated without any priori knowledge except for the lens amplification factor of a camera;
secondly, cruising by using 0.2 times of speed according to the generated PTZ track information through a constant-speed cruising function of the cradle head, and directly reading an image to be identified through a video;
feature three, recognition first stage: firstly carrying out fuzzy recognition on an image to be recognized, discarding all photos with high fuzzy rate, then carrying out real-time target detection recognition, extracting an image of a suspected disease withered tree, and simultaneously calculating corresponding PTZ information; this is the first identification of the method, which uses the YOLO V3 target detection algorithm of open source for identification;
feature four, recognition second stage: and re-detecting the PTZ information of the smaller suspicious image found in the last step, and calculating the re-identification of a new PTZ stage which needs to be positioned to be of a proper size. Cutting out images needing further classification processing;
and fifthly, performing CNN classification and identification on the cut image to further improve accuracy. The classification can greatly improve the recognition accuracy from 40% to more than 80%, and truly solves the difficulty of low target detection accuracy. If the results of the two-stage recognition are successful, outputting the result and the corresponding PTZ value, otherwise discarding.
Finally, it should be noted that: the above description is only illustrative of the specific embodiments of the invention and it is of course possible for those skilled in the art to make modifications and variations to the invention, which are deemed to be within the scope of the invention as defined in the claims and their equivalents.
Claims (4)
1. A pine wood nematode disease dead tree detection method based on a two-stage high-altitude tripod head video is characterized by comprising a first stage and a second stage, wherein the first stage is used for directly and quickly positioning a PTZ value of an appearance position of a suspicious disease dead tree by using the video; the second stage of re-photographing the suspicious disease dead tree, and adopting a pine wood nematode disease dead tree CNN classification algorithm to perform further identification treatment;
the first stage comprises a first step, a second step and a third step, and the second stage comprises a fourth step and a fifth step;
step one: generating a PTZ scanning information video which rotates horizontally by utilizing a parameter value Z of a high-altitude pan-tilt camera;
step two: for each frame image F of the video generated in step one k Performing recognition pretreatment to remove fuzzy images;
step three: performing target detection and identification on the pretreatment result of the step two by adopting a target detection pine wood nematode algorithm, and extracting suspicious image results;
step four: the image F needing repositioning shooting in the third step k Reprocessing, calling the high-altitude tripod head camera to zoom and position, and photographing to generate a new positioning mapAn image; performing recognition pretreatment of the second stage; then, target detection and identification are carried out on the pine wood nematode through a target detection algorithm;
step five: classifying and identifying by using a classified pine wood nematode dead tree CNN classification algorithm, and finding out an image of a classification result;
the first step comprises the following steps:
step 1.1: turning on PTZ display function of high altitude PTZ camera, and simultaneously reading zoom multiple parameter of high altitude PTZ camera, namely parameter value Z, and obtaining minimum value Z of parameter value Z min Maximum Z max ；
Step 1.2: triple-loop scanning is carried out on PTZ of the high-altitude PTZ camera, video is shot through the high-altitude PTZ camera, and the P value is from P min ～P Max Regulation, P min Is the minimum value of P value, P Max Is the maximum value of the P value; t value from T min ～T max Regulation, T min Is the minimum value of T max Is the minimum of the T value; z value from Z min To Z max Adjusting; the innermost loop is the P-value loop, whenever it reaches a minimum value of P min Or maximum value P Max When the T value is regulated; when the T value reaches the maximum value T max Or minimum value T min When the Z value is adjusted;
the step 1.2 comprises the following steps:
step 1.2.1: the high-altitude pan-tilt camera moves horizontally at a speed which is 1/m times the maximum horizontal movement speed, and m is more than 1; horizontally scanning and shooting video;
step 1.2.2: if the P value is greater than P min And is less than P max Returning to the step 1.2.1; when the P value is greater than or equal to P max Or less than or equal to P min When the horizontal rotation is reversed, the T value is regulated by the stepping value delta T, and delta T represents the stepping value of the high-altitude tripod head camera in the vertical direction; step 1.2.3;
step 1.2.3: if the T value is greater than T min Less than T max Returning to the step 1.2.1; if T value>T max Then T value = T max The method comprises the steps of carrying out a first treatment on the surface of the If T value<T min Then T value = T min The method comprises the steps of carrying out a first treatment on the surface of the Adjusting the Z value by Δz, Δz representing each shrinkageAdjusting multiple of the placement; turning to step 1.2.4;
step 1.2.4: if the Z value is less than Z max Returning to the step 1.2.1; if Z value is greater than or equal to Z max The first stage scanning is finished;
the third step comprises the following steps:
step 3.1: reading an image F k Image data of (2) and corresponding PTZ value, i.e. P k T k Z k Putting the target into a target pine wood nematode detection algorithm for identification, wherein the target pine wood nematode detection algorithm is a trained YOLO V3 algorithm, and identifying whether the target pine wood nematode detection algorithm is an image of dead trees or not to obtain an identification result;
then, selecting the identification result coordinates by a frame, and outputting the identification result coordinates as follows:
aBoxs＝[{x 11 ,y 11 ,x 12 ,y 12 },{x 21 ,y 21 ,x 22 ,y 22 },...,{x i1 ,y i1 ,x i2 ,y i2 },...]the method comprises the steps of carrying out a first treatment on the surface of the Wherein x is i1 ，y i1 Representation of
The upper left corner coordinate, x, corresponding to the identified dead tree image i i2 ,y i2 Representing its lower right angular position;
step 3.2: judging the current image F k Whether the image is an image of dead trees or not, if not, discarding the image F k Go to step 3.1; if yes, turning to step 3.3;
step 3.3: judging { x } i1 ,y i1 ,x i2 ,y i2 Whether or not the framed image area is smaller than image F k 1/4 of the picture of (a), if so, image F k Marked with { x } i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k 1, need to relocate the photo; otherwise, image F k Marking, namely: { x i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k 0, without repositioning the photograph, put the marked result into the mark array R1;
step 3.4: image F to be photographed without repositioning k According to { x } i1 ,y i1 ,x i2 ,y i2 Cutting the coordinate position, and putting the cutting result into an array R2;
step 3.5: judging whether all the videos generated in the step two are processed, if yes, turning to the step four, otherwise turning to the step 3.1;
the fourth step comprises the following steps:
step 4.1: sequentially reading the position information of the image information in the array R1, namely the array generated in the step 3.3
aBoxs＝[{x 11 ,y 11 ,x 12 ,y 12, P 1 ,T 1 ,Z 1 ,x 1 },{x 21 ,y 21 ,x 22 ,y 22, P 2 ,T 2 ,Z 2 ,x 2 },,...,{x i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k ,x i },...]；
Step 4.2: if { x } i1 ,y i1 ,x i2 ,y i2, P k ,T k ,Z k ,x i X in } i =1, then according to the resolution and x of the current video of the aerial pan-tilt camera i1 ,y i1 ,x i2 ,y i2 Calling a 3D positioning API of the high-altitude tripod head camera according to the coordinate x i1 -10,y i1 -10,x i2 +10,y i2 +10, scaling and positioning;
step 4.3: photographing to generate a new positioning image, carrying out target detection and identification on the new positioning image, and then, selecting the identification result coordinate { x 'again through a target detection pine wood nematode algorithm' i1 ,y’ i1 ,x’ i2 ,y’ i2 }；
Step 4.4: if the identification result is not the image of dead trees, discarding the image, and turning to the step 4.1; otherwise, the PTZ value is read again, and the output result is { x' i1 ,y’ i1 ,x’ i2 ,y’ i2, P’ k ,T’ k ,Z’ k ,0}；
Step 4.5 according to the new recognition result coordinates { x' i1 ,y’ i1 ,x’ i2 ,y’ i2 The recognition result is cut out and put into an array R2.
2. The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video of claim 1, which is characterized by comprising the following steps: the second step comprises the following steps:
step 2.1: sequentially reading each frame image F in the video of Gao Kongyun cameras k ；
Step 2.2: judging image F using OpenCV algorithm k Whether the images belong to a fuzzy image; if so, discard the image F k Turning to step 2.1; otherwise, turning to step 2.3;
step 2.3: for image F k OCR recognition is carried out to obtain an image F k In real time PTZ value, P k T k Z k Image F k And P k T k Z k The values are combined together.
3. The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video of claim 2, which is characterized by comprising the following steps: the step 2.2 includes: for image F k Performing Laplacian transformation, judging its return value, taking 100 as threshold value, if the return value is less than 100, the image F k Discarding the image F as a blurred image k Turning to step 2.1; if the return value is greater than 100, the clear diagram is considered, and the step 2.3 is shifted.
4. The pine wood nematode disease dead tree detection method based on the two-stage high-altitude tripod head video of claim 1, which is characterized by comprising the following steps: the fifth step comprises the following steps:
step 5.1: sequentially reading image information F 'in array R2' i ；
Step 5.2: for image information F' i Classifying and identifying by using a classified pine wood nematode disease dead tree CNN classification algorithm, wherein the method uses a trained VGG classification algorithm;
step 5.3: if so, outputting image information F' i Corresponding PTZ valueThe method comprises the steps of carrying out a first treatment on the surface of the If not, turning to step 5.4;
step 5.4: and judging whether the array R2 is processed completely, if not, turning to the step 5.1, and if so, ending.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110040676.3A CN112800867B (en) | 2021-01-13 | 2021-01-13 | Pine wood nematode disease dead tree detection method based on two-stage high-altitude tripod head video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110040676.3A CN112800867B (en) | 2021-01-13 | 2021-01-13 | Pine wood nematode disease dead tree detection method based on two-stage high-altitude tripod head video |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112800867A CN112800867A (en) | 2021-05-14 |
| CN112800867B true CN112800867B (en) | 2023-05-12 |
Family
ID=75810331
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110040676.3A Active CN112800867B (en) | 2021-01-13 | 2021-01-13 | Pine wood nematode disease dead tree detection method based on two-stage high-altitude tripod head video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112800867B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113905178B (en) * | 2021-10-12 | 2023-05-30 | 重庆英卡电子有限公司 | Environment automatic sensing cruising method based on high-altitude holder |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109948563A (en) * | 2019-03-22 | 2019-06-28 | 华南农业大学 | A kind of withered tree detection localization method of the pine nematode based on deep learning |
| CN110136447A (en) * | 2019-05-23 | 2019-08-16 | 杭州诚道科技股份有限公司 | Lane change of driving a vehicle detects and method for distinguishing is known in illegal lane change |
| CN110232307A (en) * | 2019-04-04 | 2019-09-13 | 中国石油大学(华东) | A kind of multi-frame joint face recognition algorithms based on unmanned plane |
| CN111611989A (en) * | 2020-05-22 | 2020-09-01 | 四川智动木牛智能科技有限公司 | Multi-target accurate positioning identification method based on autonomous robot |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9317860B2 (en) * | 2011-03-08 | 2016-04-19 | Bank Of America Corporation | Collective network of augmented reality users |
-
2021
- 2021-01-13 CN CN202110040676.3A patent/CN112800867B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109948563A (en) * | 2019-03-22 | 2019-06-28 | 华南农业大学 | A kind of withered tree detection localization method of the pine nematode based on deep learning |
| CN110232307A (en) * | 2019-04-04 | 2019-09-13 | 中国石油大学(华东) | A kind of multi-frame joint face recognition algorithms based on unmanned plane |
| CN110136447A (en) * | 2019-05-23 | 2019-08-16 | 杭州诚道科技股份有限公司 | Lane change of driving a vehicle detects and method for distinguishing is known in illegal lane change |
| CN111611989A (en) * | 2020-05-22 | 2020-09-01 | 四川智动木牛智能科技有限公司 | Multi-target accurate positioning identification method based on autonomous robot |
Non-Patent Citations (6)
| Title |
|---|
| A deep convolutional neural network for video sequence background subtraction;Mohammadreza Babaee;《Pattern Recognition》;635-649 * |
| Automatic PTZ Camera Control Based on Deep-Q Network in Video Surveillance System;Dongchil Kim;《2019 International Conference on Electronics, Information, and Communication (ICEIC)》;1-10 * |
| 基于Faster R-CNN的松材线虫病受害木识别与定位;徐信罗;陶欢;李存军;程成;郭杭;周静平;;农业机械学报(第07期);1-5 * |
| 基于FreeRTOS的嵌入式云台控制系统设计;朱耀麟;《2015年电子技术应用》;1-5 * |
| 多PTZ主动摄像头的类目标检测定位系统;马庆平;《中国优秀硕士学位论文全文数据库 (信息科技辑)》;I138-6468 * |
| 浅谈PTZ摄像机的技术应用与发展方向;尹雪辉;《中国公共安全》;1-6 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112800867A (en) | 2021-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11765321B2 (en) | Intelligent video surveillance system and method | |
| CN104754302B (en) | A kind of target detection tracking method based on rifle ball linked system | |
| CN105894702B (en) | A kind of intrusion detection warning system and its detection method based on multiple-camera data fusion | |
| CN111291633B (en) | Real-time pedestrian re-identification method and device | |
| CN101918989B (en) | Video surveillance system with object tracking and retrieval | |
| KR101781358B1 (en) | Personal Identification System And Method By Face Recognition In Digital Image | |
| WO2020151750A1 (en) | Image processing method and device | |
| US20040183679A1 (en) | Thermal signature intensity alarmer | |
| TWI679612B (en) | Image tracking method | |
| KR100326203B1 (en) | Method and apparatus for face photographing and recognizing by automatic trading a skin color and motion | |
| CN108810413A (en) | Image processing method and device, electronic equipment, computer readable storage medium | |
| CN110493517A (en) | The auxiliary shooting method and image capture apparatus of image capture apparatus | |
| CN111163285A (en) | High-altitude falling object monitoring method and system and computer readable storage medium | |
| CN107181918A (en) | A kind of dynamic filming control method and system for catching video camera of optics | |
| JP4764172B2 (en) | Method for detecting moving object candidate by image processing, moving object detecting method for detecting moving object from moving object candidate, moving object detecting apparatus, and moving object detecting program | |
| CN107609475B (en) | Pedestrian detection false detection extraction method based on light field camera | |
| CN112800867B (en) | Pine wood nematode disease dead tree detection method based on two-stage high-altitude tripod head video | |
| CN107622497A (en) | Image cropping method, apparatus, computer-readable recording medium and computer equipment | |
| CN106599779A (en) | Human ear recognition method | |
| CN113689458B (en) | 2D shooting track path calculation method and device | |
| CN109862263A (en) | A kind of mobile Automatic Target Tracking method based on the identification of image multi-dimensional feature | |
| TW466452B (en) | Method and apparatus for photographing/recognizing a face | |
| CN111199203A (en) | Motion capture method and system based on handheld device | |
| MathuraBai et al. | Object Detetcion using SSD-MobileNet | |
| Luo et al. | Image de-fencing based on binary morphology |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |